This project will study changes in (a) health care cost and utilization, (b) worker's compensation and disability payments, and (c) rates of injury for alcoholics and their family members following initiation of alcoholism treatment. A longitudinal research design with multiple pretests and posttests will be utilized to conduct a 12-year (1974-85) study of employees (and insured family members) of Deere and Company (John Deere). Study groups include (1) an experimental group of 1,000 treated alcoholics, (2) two control groups of untreated alcoholics (each N=400), matched with the experimental group on key variables, and (3) a matched comparison group of nonalcoholics. Research hypotheses concern statistically significant differences in health care costs (and other variables) of the treated alcoholic group before and after alcoholism treatment compared with the matched control groups. It is hypothesized that the treated alcoholic group will converge toward the patterns of the nonalcoholic group over time. Control and comparison groups will be matched with the treated alcoholic group by age, sex and employment location. All data will be completely anonymous and no individuals or families will be identifiable. While studies have examined the unemployed or institutionalized alcoholic, studies of employed alcoholics have been fewer and most suffer from major methodological limitations. As a result, employers and health insurance carriers are often reluctant to specifically include coverage of alcoholism treatment for fear of increased costs. Project results can make a significant contribution to our understanding of the health care costs and utilization, injury rates and worker's compensation payments of employed alcoholics and their families, as well as provide policy relevant data for health care cost containment for industry, government, and insurance companies which are derived from a natural setting. Specific features added by this supplemental proposal will increase the robustness, scope and statistical power of the project by (1) doubling the size of the study groups, (2) adding a second untreated alcoholic control group, (3) adding four years of data, (4) adding additional dependent measures, (5) adding an analysis of subpopulations, and (6) adding a time-series analysis to corroborate results of pre/post analysis.